Forecasting a region’s travel has long relied on models of human travel behavior capable of representing the interrelationships among a multitude of activity-travel and location choices that people make over time and space. However, an issue that is often overlooked in choice modeling efforts is that key phenomena underlying model specification and estimation are often unobserved. As a result, analysts end up making very significant assumptions regarding underlying choice processes that may not necessarily be reflective of the behavioral heterogeneity present in the sample. Unfortunately, in many modeling contexts, these underlying phenomena are unobserved. This presentation articulated the problem and offered two contexts where the notion of latency was explicitly incorporated into the econometric behavioral model formulation.